INTRODUCTION: Chronic inflammation is associated with increased red blood cell (RBC) alloimmunization. Patients with sickle cell disease (SCD), systemic lupus erythematosus and inflammatory bowel disease, all conditions with increased levels of inflammation, have been shown to be more alloimmunized when compared to the general population. High prevalence rates of RBC alloimmunization have also been observed in patients with myelodysplastic syndrome (MDS). It has recently been shown that an inflammatory microenvironment may be a key driver of MDS pathogenesis. There are a few specific genes recurrently mutated in MDS that have been found to be associated with activation of the innate immune system and formation of the inflammasome, including some involved in epigenetic modification - ASXL1, DNMT3A, EZH2 and TET2 - and spliceosome formation - SF3B1, SRSF2 and U2AF1. The aim of this study was to determine if the genetic mutations driving the inflammatory milieu in MDS could simultaneously be creating an environment that promotes alloimmunization. PATIENTS AND METHODS: We identified patients diagnosed with a myeloid neoplasm, including patients with MDS and acute myeloid leukemia (AML), between 2015 and 2020 on whom a type and screen had been ordered. We recorded the patients’ primary diagnosis, sex, race/ethnicity, age at first sequencing, somatic mutations identified by next generation sequencing (NGS), blood type, antibody screen results, specific alloantibodies and autoantibodies identified, and number of RBC and platelet units transfused. Patients with an autoantibody, a clinically significant alloantibody, or non-specific reactivity with a potential antibody under investigation on their screen were put in the positive antibody screen category, and those with reactivity solely due to Rh immune globulin or monoclonal antibody therapy were put in the negative antibody screen category. Univariate logistic regression was run separately for patients initially diagnosed with AML and those initially diagnosed with a non-AML myeloid neoplasm (MN), using sex, age, race/ethnicity, number of RBC units transfused, and somatic mutation, to determine if any of these variables had a significant association with a positive antibody screen. If quasi-complete separation issues existed, Firth's method for separation was applied to the logistic regression model. All analyses were conducted using a 95% confidence level. RESULTS: Seventy-six patients met inclusion criteria; 44 were male and 32 were female, and ranged in age from 14 to 89 years. The majority of patients were Caucasian (n = 64). Thirty-three patients had an initial diagnosis of AML; 43 were initially diagnosed with a MN, including MDS (n = 36), myeloproliferative neoplasm (n = 2), chronic myelomonocytic leukemia (n = 4) and chronic myeloid leukemia (n = 1). Number of RBC units transfused ranged from 3 - 112 in patients with AML and 0 - 131 in patients with a MN. Thirteen patients with AML had a positive antibody screen (39.4%), compared to 21 patients with a MN (48.8%). Patients with AML most often developed an alloantibody without an accompanying autoantibody (58.3%), while patients with a MN most often developed alloantibodies in addition to autoantibodies (52.4%). Patients with a MN and a SRSF2 mutation had a significantly increased rate of allo- or auto-immunization as compared to patients with AML and a SRSF2 mutation (p-value = 0.035). Similarly, patients with a MN and a TET2 mutation had a significantly increased rate of allo- or auto-immunization as compared to patients with AML and a TET2 mutation (p-value = 0.021). CONCLUSIONS: In patients with a MN, the presence of a SRSF2 or TET2 mutation predicted the presence of at least one allo- or autoantibody. Our data support the hypothesis that genetic mutations involved in formation of the inflammasome may be involved in the increased alloimmunization seen in patients with MDS. Our study of alloimmunization in MDS, and further studies like it, may enable us to predict based on genetic mutations which patients with MDS may be at increased risk for becoming alloimmunized. Similar to prophylactic C, E and K antigen matching that some blood banks employ in patients with SCD to minimize formation of alloantibodies, extended phenotype matching for the most immunogenic antigens could also be considered in patients with MDS and SRSF2 or TET2 mutations to decrease alloimmunization in this patient population.
Read full abstract